ARM Predicts First Cortex-A15 Devices Late 2012

Interesting.. I wonder what the use of a 16 core mobile platform would be, lol.

intel will molest arm and take over the world INTEL FTW MOFOS!!!!!!!!!!

It's not necessarily used in a mobile platform.

"and keyboards and mice will one day connect via Bluetooth"

huh? you can get them now, they have been available for years...

However i threw out my bluetooth keyboard/mouse. Wireless sucks, replacing batteries all the time sucks.

------

Or do you mean connecting to a phone.... WTF would you lug around a keyboard, mouse and dock to turn a phone into a laptop.... and dont forget the screen.... Just buy a laptop and have a better platform.

Phones aren't and will never be a laptop. They work ok for content consumption but not creation. And they never will work for content creation....unless you lug around a dock, a keyboard, a mouse, and a lcd....

That is unless we figure out how to make some kind of projected display/interface that isnt garbage. (ie mass effect style) If we crack that nut....sure...then it can replace a laptop.

...Obviously. The A15, however, was designed with a mobile platform in mind, hence the relevance of my statement. A non-mobile platform, while possible, is not what they focused on designing it for, therefore, one must reasonably conclude that when they designed this mobile-oriented chip.. that adding scalability up to 16 cores.. was part of their eventual mobile platform vision. Unless, of course, they publicly say otherwise. Which they haven't.. Which, brings us full circle, back to my statement.

I wonder what the use of a 16 core mobile platform would be, lol.

PC, laptop, netbook, tablet, smartphone would most likely merged into one single portable device in the future. Bascially, if we're at home, we could dock the smartphone onto a full blown monitor, keyboard, mouse setup. Or if we're on the go, it would be just a small touch screen phone that fits inside our pocket or holster.

As with the 16 multi core on a smartphone, simple... if you're running on AC, then all cores are enabled (home use). And if you're running on battery, then only a few cores would be enabled (on the go use).

I highly doubt they will go higher than 4 cores in mobile devices anytime soon.

The fact that Intel can't beat ARM on the same node means that they'll never beat ARM. It's not like Intel is magically capable of reaching nodes that nobody else will ever reach. Intel can only succeed where they can use their IP and monopoly influence to shut everyone else out, aka x86 desktops. ARM will probably begin cannibalizing x86 laptops in a major way soon.

Agreed. Very much so. I'm just doing a little "daydreaming" really. I'm struggling to see a real use for a 16 core mobile device.. However, as I would have said the same about a mobile 2GHz+ 2/4 core 10 years ago.. It obviously doesn't mean that much really. But curiosity is still curiosity.

ARM is a great company to say the least.

And the OS for tablets and mobile phones is good enough for many users that just want to push and play. Featuring keyboard, mouse and other input types will make these platforms the choice for most users, not the power users but the masses will choose something like that instead (think parents and grandparents). That's why Intel is trying to make Android for x86. That's why Windows 8 will be made with mobility in mind.
The appearance of the OS with mobility like features will not prevent heavy productive work or intensive entertainment from being done.
So, here, aiming for the future, all of them, ARM, Intel, Microsoft, Nvidia, Google. Making people have a familiar interface and stop thinking about the rest if they wish not to think about it. Power tools and power users will always exist. Most users, though, will have a docking station that has wireless everything (including inductive charging), and they put the phone there, 1-2 cores, the OS of the phone will synchronize to the OS of the docking station and the same workspace will go on the other device, even as simple as keyboard and screen, or as complex as HDTV and Kinect. The data would be mostly on the cloud anyway, only the local rendering will need the power, so, the more power, the more details and resolution and framerate, but with all the portability needed. Even streaming of pre-rendered video for high demanding entertainment, on the phone, with an intermediate step of being rendered almost locally, for example in the docking station, and then streamed to the phone. All calculations can't be done remotely because of the delay in response time. The delay is tolerable in a locally rendered environment like WoW, because the appearance is "as if you were here", with environment settings being here already, and movements being predicted somehow and synchronized and fluid enough. If everything is rendered in the cloud and streamed over, the experience would never match.
This is the future as I see it from the data at hand. More power, and more mobility, with the dark cloud looming over us all. Not a fan of the cloud, but recognizing it.

Shadowrun. Decks.

I can see adoption of this chip in the server markets a potential issues since it is only 32-bit. I work for a Fortune 100 company within the Network Operations department and we order only a few different model of servers for all of our server types(four to be exact: 1x type of a Blade Dual Socket, 1x type of a Blade Quad Socket, 1x type of a Dual Socket traditional, 1x type of a Quad Socket traditional). Even if these are more power efficient, we won't be using them any time soon since we like to keep all of our servers as close to the same configuration as possible. This simplifies troubleshooting issues along with making it easier to explain to the business side that we offer only four server models, so no time wasted in deciding, testing, etc what server is best.

On top of that, most of our new servers are being built as VMs under VMWare vSphere. You can't mix and match different processor types within a cluster due to live migration issues... as in if a VM is running on a Host running an Intel processor, it can only migrate to another Host running an Intel processor since the OS is using extensions based on that type of processor. With 32-bit and 64-bit VMs running on the same Hosts, it makes it much easier to manage and wouldn't want to think of what it would take to manage multiple farms based on different Architecture types.

I understand all of the above is based on an OS built for the x86 architecture, but this would introduce another headache...support for another version of a server OS built for ARM. To much support costs in integrating a new architecture that is only 32-bit, especially when most everything is moving towards 64-bit. If this chip supported 64-bit as well as 32-bit, then support costs could be cheaper since you could then migrate everything over if you wanted too(assuming that the OS and Apps all supported ARM).

"... and keyboards and mice will one day connect via Bluetooth..."

They already should be able to. The problem is, at least for the Droid 2, is that the Bluetooth HID implementation is poor (HID is the protocol used by mice, keyboards, and similar devices). When the mouse or keyboard temporarily closes the Bluetooth connection to go into low power mode (ex: user hasn't typed in 20 minutes) and then later wakes up and tries to reconnect to the phone, the phone does not allow the device to reconnect. Thus, you can pair your Bluetooth mouse or keyboard and it will work the first time for as long as you are actively using the device, but if you walk away or turn one of the devices off, the device stops working and you are forced to either re-pair again or give up. Considering that all the hard work of support Bluetooth HID devices has already been done (i.e. converting data packets from the device into keystokes or mouse movements on the screen), this is a painfully lame bug.

Note: I have not re-tested this with the latest Droid 2 system update.